Position Paper for CHI'95 Research Symposium (May 6-7, 1995, Denver, Colorado U.S.A)
User Participation in Task-based Design Stephanie Wilson, Peter Johnson and Hilary Johnson Department of Computer Science Queen Mary and Westfield College Mile End Road London E1 4NS, UK Tel: 44-71-975-5231 E-mail: steph, pete,
[email protected] ABSTRACT
This position paper addresses the participatory design theme of the research symposium. Previous research on taskbased design has provided the basis for a current project concerned with empowering users in the design process. We describe early work on this project where we hope to investigate how users might participate in formulating descriptions of their work contexts and in deriving artifact designs which will have consequences for those contexts. KEYWORDS: Participatory design, task-based design BACKGROUND
Earlier work [2] [3] [5] reported research into a task-based approach to interactive system design and examined how representations of users' work tasks could form the basis for the design of interactive systems. The design approach was supported by software known as the ADEPT toolset. The motivation behind our work on task-based design was to enable designers to produce interactive system designs that are fit for their purpose, insomuch as the system is compatible with the tasks it is intended to support. (A task, in this context, is a task of the user's work, not a lowlevel interaction task.) We explored three distinct aspects of task-based design in ADEPT, namely the design process, design representations (notations) that might be used at different stages of the design process, and design support tools. The design process started with a task analysis procedure and then continued with designers expressing the task data as a task model, modelling proposed changes to users' tasks, and integrating this knowledge into system design. ADEPT might be said to exemplify a 'traditional' approach to HCI. Although the design method emphasised a highly iterative process, involving both top-down and bottom-up activity, it was still focused around a conventional analysisdesign-evaluation cycle. We were concerned with supporting the designer's, rather than the user's, role in the design process, leaving user involvement still at the stage of evaluating designs and design representations, as opposed to participating in their creation. Furthermore, the ADEPT tools were developed to support the user interface designer, and were not intended at the outset to support user participation in design. However, experience with ADEPT sug-
gested that tools such as the task model editor might prove a valuable way of obtaining user feedback and getting users directly involved in modifying the design representations. CURRENT RESEARCH DIRECTIONS
In the so-called 'Scandinavian school' of participatory design, users act as fully empowered participants in the design process. They are given the opportunity to affect decisions that will later impact their work lives. Our current research on interactive system design is influenced by this approach. Given that our design method already focuses on modelling users' work, proposed changes to that work, and on developing designs to support work, it is a natural next step to investigate the involvement of users in the modelling process, rather than simply as subjects and evaluators. Our primary motivation in taking this step is to increase user participation in design with a view to enhancing the usability of artifacts, rather than for reasons of any philosophy of industrial democracy. Work context: users, organisations, existing artifacts, etc task modelling
consequences for work context
Existing task descriptions
requirements gathering
Requirements
design
Proposed task descriptions design
Artifact designs
Figure 1: Task-based design
As before, we are concerned with a task-based design process in which the users' current tasks and proposed changes to those tasks are explicitly represented. The difference is that users, as members of the design team, are empowered to participate in the design activities. In the traditional ap-
Position Paper for CHI'95 Research Symposium (May 6-7, 1995, Denver, Colorado U.S.A) proach there are problems for designers in understanding and representing the current work context and there are problems for users in assessing what consequences technical solutions will have for their future work. We hope to address these issues through active user participation and direct engagement of the designers with the users. Figure 1 provides a simplified overview of the design process, showing it as a purely top-down activity (which is clearly not the case). Design begins with the team developing a common understanding, through joint modelling, of the current work context, the design problem space, and the requirements for changes to the work. Requirements generally provide the impetus for introduction of new technology or artifacts. They determine that people's work should be changed in some manner, and influence the development of new artifacts to support the new work. Design methods often focus on the design of the new artifact but disregard the real change: the design of new work. Our approach suggests that the design team should consider the proposed changes to work (the 'Proposed task descriptions'), prior to exploring detailed designs (the 'Artifact designs') for systems to support the new work. The introduction of new artifacts and the changes to people's work have consequences for the entire work context (Figure 2).
work context work tasks
Introduction of new design
new work context new work tasks
are exploring is greater use of task scenarios and user interaction scenarios in which both designers and users participate. In addition, while we recognise that software tools may be an obstacle, rather than a solution, to the question of how to increase user participation, the task modelling notations used in ADEPT proved to be useful (and comprehensible) by both designers and users. We are presently investigating these further and looking to see whether improved implementations of our task-modelling tools could be employed by the user community. SUMMARY
This position paper has outlined the direction of our current research into increasing user participation within a taskbased approach to design. Design studies, to be carried out in the near future, will help to assess the appropriateness of various techniques for promoting user participation in a task-based design process. A second aspect to the design studies will be to acquire additional information about existing design practice, with a view to broadening the scope of our approach to encompass aspects of work context beyond user tasks. Moreover, we recognise the potential value of established mechanisms for carrying out design and the inherent difficulties in changing design practice, even where such changes are well motivated. A related question that must also be addressed concerns the extent to which full participatory design, with users as equal participants, is a realistic proposition in the context of the UK workplace. ACKNOWLEDGEMENTS
The work reported here is supported by the UK Engineering and Physical Sciences Research Council (EPSRC). REFERENCES
Figure 2: Introduction of new designs and the work context
The first phase of our research is concerned with investigating techniques to increase user-designer communication in this task-based design process. For a start, the task modelling process ceases to be a process of analysis (and subsequent modelling) by the human factors expert, with only limited input from the users, and equally limited validation of the resulting model by the users once it is complete. Instead, it becomes a process in which both users and designers can participate to model existing work-task situations, to derive task-oriented design requirements and to model proposed new work-task situations. Other position papers submitted to this forum examine user participation in the particular form of task analysis (TKS [1]) employed in ADEPT, and discuss the role of design representations. Likewise, we aim to investigate possibilities for greater user participation in the actual design of artifacts. Initially, we propose to assess the suitability of existing design techniques for application at different stages in design. These include the graphical diagrams used in ADEPT, ethnomethodology techniques, hypertext recording and PICTIVE-like design objects [4]. Another approach we
1. Johnson, H. and Johnson, P. Task knowledge structures: Psychological basis and integration into system design. Acta Psychologica, 78, 1991, pp. 3-26. 2. Johnson, P., Johnson, H. and Wilson, S. Rapid Prototyping of User Interfaces Driven by Task Models. J. Carroll (ed), Scenario-based design, John Wiley & Sons, Inc (In Press). 3. Johnson, P., Wilson, S., Markopoulos, P. and Pycock, J., ADEPT—Advanced Design Environment for Prototyping with Task Models, Demonstration Abstract. In Proceedings, InterCHI'93, ACM Press, 1993, p. 56. 4. Muller, M. PICTIVE: democratizing the dynamics of the design session. In D. Schuler and A. Namioka (eds), Participatory design: principles and practices, Lawrence Erlbaum, N.J., 1993. 5. Wilson, S., Johnson, P., Kelly, C., Cunningham, J. and Markopoulos, P. Beyond hacking: a model based approach to user interface design. In Proceedings of HCI'93, J. Alty, D. Diaper and S. Guest (eds), Cambridge University Press, 1993, pp. 217- 231.